Classic editor came first. A textarea, a few buttons, raw HTML if you wanted it. It worked because the people building sites were mostly the same people who knew code.
Then Gutenberg landed in 2018. Blocks, a visual hierarchy, a different mental model. The shift wasn't subtle, and the pushback was loud. People still write Classic-Editor-only plugins. People still argue about it. But the direction was clear: more abstraction, less HTML.
Page builders followed the same arc. Elementor, Divi, Bricks. Each one moved the user further from code and closer to outcomes. Each one was called a hack, a workaround, a sign that core had failed. Each one is now used by millions of sites.
Three UI shifts. Each larger than the last. Each one resisted by power users who saw the previous layer as fine, even superior. I think we're in the fourth shift now, and most of the WordPress community is missing what's actually happening.
The shift: AI isn't a tool you bolt onto WordPress, it's the next interface layer
For the past two years, “AI in WordPress” has mostly meant a button that wrote a paragraph for you. A plugin that generated alt text. A widget in the admin bar. Useful, but additive. The block editor was still the editor. The AI was a feature inside it.
That's not what's coming.
What's coming is conversation as the primary interface. You describe the change, the system makes it. Not a paragraph, not alt text, the actual page. Layout, content, images, structure. You describe the bakery, you get the bakery homepage. You describe the typography you want, you get it across the site. This is technically possible now in a way it wasn't twelve months ago. Not because the models got bigger but because they got good enough to operate on real Gutenberg structures, real REST API calls, real custom fields, real hooks. The semantic gap between “user intent” and “WordPress action” closed. That gap was the bottleneck. It's gone.
Why WordPress, specifically, wins this round
The argument I want to make is that WordPress, specifically, is the platform best positioned for this layer. Not Wix. Not Squarespace. Not Webflow.
Open architecture is the reason.
A real co-pilot has to touch everything. Posts, pages, blocks, taxonomies, custom fields, theme settings, plugin behaviour. It has to read the actual page, understand the actual context, and write back through the same channels a human would. Hooks, filters, the REST API, WP-CLI, the database when needed.
Closed platforms don't expose this surface. They can build an AI layer on top of their UI, but the AI is constrained to whatever buttons the UI has. WordPress isn't. A co-pilot inside WordPress can do anything a developer could do, because the developer surface is fully open.
This is why the Wix and Squarespace AI demos always feel a bit small. They generate a starter site, then hand you back to the same builder you would have used anyway. The AI is a wizard at the front, not a layer through the whole stack.
What this means for WordPress professionals
I want to address the fear directly: “if AI builds the site, what's left for me?” A few things, all important.
First, AI co-pilots don't replace expertise. They replace effort. When a model can generate ten variations of a homepage in two minutes, the bottleneck moves from “can I build this” to “which version is right for the brand, the audience, the conversion goal.” That's craft. That's judgement. That's exactly what experienced WordPress people are good at.
Second, the lower end of the market enters WordPress that wouldn't have. Right now, a small business owner who can't afford a developer goes to Wix. They were never going to call you. With a conversational layer, they can start on WordPress, scale on WordPress, and call you when they need real work done. The funnel into the WordPress ecosystem widens, not narrows.
Third, plugins and themes don't go away. They become more important. The AI needs context to do good work. A site running ACF, a custom post type, a translation plugin, an analytics integration: all of that gives the AI the structure it needs to make meaningful changes. Plugins are the language the AI speaks.
What stays
This is the part that gets lost in the AI panic. WordPress, the platform, doesn't change. wp-content stays. The database stays. The Loop stays. Plugins, themes, hooks, filters, custom fields, REST API. All of it.
What changes is the entry point. You can still write your post in the block editor if that's what you want. You can still hand-code a custom theme. You can still install a page builder. The new layer doesn't take anything away. It adds an option.
This is not how AI is being framed in most platforms. Most platforms position AI as a replacement: “skip the editor, just describe what you want.” WordPress can do something subtler. It can make AI a colleague that uses the same tools you do, and hands the work back to you when you need to take over. That's a much more interesting product, and it's only possible because WordPress is open.
Closing
The platform that internalises this layer first wins the next five years. The mass-market user wants outcomes, not editors. They will go wherever the friction is lowest. Right now that's Wix and Squarespace, because those platforms have invested in the front-door experience while WordPress has been arguing about block patterns.
WordPress has every ingredient to win this. Open architecture. The largest plugin ecosystem on earth. 43% of the web. A community that has navigated three UI shifts already. What it doesn't have, yet, is consensus that this is the moment. That the conversational layer is real, is here, and is going to define how non-technical people build websites for the next decade. I think it is. I think WordPress was built for it. And I think the people in this WordPress community (aff), the ones who've shipped through Classic, Gutenberg, and the page builders, are the ones who get to define what comes next.
